The NIH mouse potency test for rabies vaccine is a highly variable assay system. We have been attempting to replace this test with a more reliable in vitro system such as the Single Radial Immunodiffusion Test (SRID) and the antibody binding test. The SRID test seems to be the most promising candidate at this time. We currently have three licensed manufacturers of rabies vaccine in the U.S. One product is adsorbed on aluminum phosphate and the other two are unabsorbed. We are testing each lot of the latter product by SRID and approximately 60% of the lots by the NIH test. The manufacturer is also testing each lot by both methods and reporting these results to us so that a statistical comparison of the two methods can be performed. The adsorbed vaccine cannot be tested by SRID, therefore, the antibody-binding test (using BHK-21 cells) and FITC-conjugated antirabies serum is used for this product in parallel with the NIH test. We also participated in a WHO collaborative study comparing the above test methods in the replacement of the International Standard Rabies Vaccine. A report for this study was presented to the WHO Expert Committee on Biological Standardization and was published in Biologicals (vol 20, pp 301-313) in 1992. The conclusion from this report was that the immunogenicity test (NIH) cannot be replaced by the antigenicity test (SRID, AbB, etc.) at this time. We will continue to use SRID as a screening method with U.S. licensed rabies vaccines and explore ways to achieve better correlation between NIH and SRID tests.